Mutagenesis vol. 27 no. 3 pp. 323–327, 2012 doi:10.1093/mutage/ger082 Advance Access Publication 15 November 2011 DNA replication occurs in all lamina positive micronuclei, but never in lamina negative micronuclei

Atsushi Okamoto, Koh-ichi Utani1 and Noriaki Shimizu* broken chromatin bridge connecting the separating anaphase Department of Biofunctional Science and Technology, Graduate School of . Furthermore, micronuclei can be generated from Biosphere Science, Hiroshima University, aggregates of extrachromosomal elements such as double Higashi-hiroshima, Hiroshima 739-8521, Japan, minutes (DMs) chromatin bodies, which is a cytogenetic 1Present address: Laboratory of Molecular Pharmacology, National

manifestation of gene amplification (3). This type of micro- Downloaded from https://academic.oup.com/mutage/article/27/3/323/1055360 by guest on 28 September 2021 Institute, National Institute of Health, Bethesda, MD 20892, USA. nuclei is referred to as DM-type, while the other micronuclei *To whom correspondence should be addressed. Tel: þ81-824-24-6528; are known as chromosome type (4). In addition to post-mitotic Fax: þ81-824-24-0759; Email: [email protected] generation, micronuclei are also formed during by Received on September 3, 2011; revised on October 7, 2011; nuclear budding (5). The nuclear buds may form through the accepted on October 17, 2011 lamina break, which is coupled to cytoplasmic membrane blebbing (Koh-ichi Utani, Atsushi Okamoto and Noriaki A is a small nucleus-like structure found in the Shimizu, submitted for publication). On the other hand, there cytoplasm of dividing cells that suffered from genotoxic stress. are micronuclei that lack the lamin B protein, a constituent of It is generally hypothesised that micronuclei content is nuclear lamina, as well as others that contain lamin B. It has eventually lost from cells, though the mechanism of been suggested that micronuclei lacking lamin B are generated how this occurs is unknown. If DNA located within the by broken chromatin bridges (6) and/or by nuclear budding micronucleus is not replicated, it may explain the loss of through the lamina break (Koh-ichi Utani, Atsushi Okamoto and micronuclei content. Because there had been no compelling Noriaki Shimizu, submitted for publication). evidence for this issue, we have addressed whether DNA It is commonly hypothesised that the content of micronuclei is located within the micronucleus is replicated this issue. Pulse ultimately eliminated from cells, though the mechanism of how labelling of bromodeoxyuridine revealed that DNA synthesis this occurs remains unknown. The micronuclei may be degraded takes place in a portion of micronuclei that contain nuclear # in situ in cytoplasm (7). Alternatively, they may be extruded out of lamin B protein. By using iodine 3 -deoxyuridine/chloro- cells since extracellular micronuclei have been detected in culture deoxyuridine double labelling, we found that all micronuclei fluid (8). Another possibility is that no DNA replication occurs in containing lamin B are replicated during one cycle, micronuclei, leading to the dilution of the unreplicated micronuclei whereas micronuclei lacking lamin B are never replicated. DNA during cell division. However, previous studies using short- This result suggests that the content of lamin B-negative term bromodeoxyuridine (BrdU) pulse labelling revealed that at micronuclei is lost during cell division. Furthermore, we least a portion of micronuclei underwent DNA synthesis, though simultaneously visualised sites of DNA synthesis, lamin B there were also micronuclei that did not incorporate BrdU (5,9,10). and the extrachromosomal double minutes chromatin, Since DNA replication inside micronuclei has not been well which contain amplified oncogenes. We found that studied (reviewed in ref. 7), we used several modern in situ although the replication timing of double minutes was techniques to examine this process more extensively. generally preserved in micronuclei, at times it differed greatly from the timing in the nucleus, which may perturb the expression of the amplified oncogenes. Taken together, Materials and methods these findings uncovered the DNA replication occurring inside micronuclei. Cell and cell culture Human colorectal tumour COLO 320DM (CCL 220) cells (11) and human cervical tumour HeLa cells (6) were obtained and maintained as described previously. For the experiments shown in Figure 2, HeLa cells at the mitotic phase were collected by the mitotic shake-off. The cells were cultured for 14 h in normal conditions and then chlorodeoxyuridine (CldU) and Introduction (4 lg/ml; Sigma, St Louis, MO, USA) were added to the culture.

The micronucleus is a small nucleus-like structure that forms Cytochemical procedure within cytoplasm, separately from the main nucleus (1,2). It BrdU (Sigma), iodine 3#-deoxyuridine (IdU) (Sigma) or CldU (ICN pharmaceut- can be detected in almost all growing mammalian cells in icals, Inc., Irvine, CA, USA) is a halogenated thymidine analogue, and it was used response to a wide variety of stimuli that cause genome to label the sites of DNA synthesis. Indirect immunofluorescence detection of damage. Therefore, the presence of micronuclei is widely used BrdU or the simultaneous detection of IdU and CldU was performed using and the protocols described in our previous paper (12). The as a hallmark of genotoxic stress. Micronuclei are generated incorporated CldU was detected with a monoclonal rat anti-BrdU (6 by several different mechanisms. They can be generated post- lg/ml; OBT0030; Oxford Biotechnology), and IdU was detected with a mono- mitotically from the lagging chromatid, and such ‘chromatin clonal mouse anti-BrdU antibody (1.5 lg/ml, cat. no. 347580; Becton Dickinson). laggards’ can be derived from (i) an acentric chromosome The former antibody binds to both BrdU and CldU but not to IdU, while the latter fragment generated by drugs or radiation that induce DNA antibody strongly binds to both BrdU and IdU and weakly to CldU. Therefore, samples were first incubated with the rat anti-BrdU antibody for 1 h at 37°Cto breaks, (ii) a whole chromosome that was not bound to detect CldU and then incubated with the mouse anti-BrdU antibody for 1 h at 37°C or merotelically bound to the spindle microtubule or (iii) a to detect IdU. These primary antibodies were detected by Alexa Fluor

Ó The Author 2011. Published by Oxford University Press on behalf of the UK Environmental Mutagen Society. All rights reserved. For permissions, please e-mail: [email protected]. 323 A. Okamoto et al.

647-conjugated donkey anti-mouse IgG (Invitrogen, Co.) and Alexa Fluor 488- cultures with BrdU for 2 h to label the DNA replication sites, conjugated donkey anti-rat IgG (Invitrogen, Co.). fixed the cells and visualised the BrdU-labelled DNA and the In the experiment shown in Figures 3 and 4, DNA synthesis was detected by EdU (5-ethynyl-2#-deoxyuridine) labelling. Because fluorescent detection of nuclear lamin B protein simultaneously. Representative images EdU employs only a chemical reaction, it enabled the simultaneous detection of of HeLa cells are shown in Figure 1A–F. We observed both lamin B by immunofluorescence and of DMs by fluorescence in situ lamin B-positive and negative micronuclei. Presence of both hybridization, both of which require multiple antibody reactions. Labelling type was reported for chromosome-type (6,14–16) or DM- and detection of EdU was performed as described previously (13). In short, type micronuclei (10,16). While we were able to detect DNA EdU (Invitrogen) was added to the culture to a final concentration of 10 lM and incubated for the indicated times. After labelling EdU, the cells were harvested, replication in some lamin B-positive micronuclei (Figure 1A and washed with phosphate-buffered saline (PBS) and fixed with 3% paraformal- B), we did not detect it in another portion of lamin B-positive dehyde (PFA) for 10 min at room temperature. Then, the cells were micronuclei (Figure 1C and D) nor in any lamin B-negative cytocentrifuged onto poly L-lysine-coated slides. The slides were reacted with 1 micronuclei (Figure 1E and F), despite the fact that DNA mM CuSO4, 100 mM ascorbic acid and 10 lM Alexa Fluor 647 azide (Invitrogen) for 30 min at room temperature and were washed with PBS. Denaturation of replication occurred within the main nucleus. To address DNA, hybridisation with DIG-labelled probe prepared from c-myc cosmid DNA, whether this was due to cells being in different stages, and simultaneous detection of the hybridised probe and the lamin B protein was we classified the cells into early, middle and late S phase as Downloaded from https://academic.oup.com/mutage/article/27/3/323/1055360 by guest on 28 September 2021 performed using the same procedure as described previously (10). well as G1/G2 phase according to the nuclear distribution Immunofluorescent detection of lamin B protein used 5 lg/ml Goat anti Lamin B (M-20; Santa Crutz Biotechnology, Inc.) and 10 lg/ml Texas red- conjugated rabbit anti-goat IgG (EY Laboratories).

Microscopy The images appearing in Figures 1 and 2 were obtained with an Olympus FV10-ASW confocal system on FV1000D-IX81 with a 60 objective (UPLSAPO NA 1.35, oil). The images appearing in Figure 3 were obtained with a Nikon inverted microscope (TE2000-E; Nikon, Tokyo) with a 60 objective (Nikon, Tokyo, Plan Apo VC 1.40, oil). All images were processed and assembled using Adobe Photoshop CS version 8.0.1 (Adobe Systems Inc.).

Results and discussion BrdU pulse labelling revealed that DNA replication occurs in lamin B-positive micronuclei About 5% of logarithmically growing cultures of cells (HeLa or COLO 320DM) contained micronuclei. We treated these

Fig. 2. DNA replication occurs in all lamin B-positive micronuclei but not in lamin B-negative micronuclei. (A) Outline of the experimental procedure. Logarithmically growing COLO 320DM or cell cycle-synchronised HeLa cells were cultured in the presence of 30 lM CldU and 4 lg/ml cytochalasin B for 12 h. Then, 100 lM IdU was added during the last 20 min. The cells were Fig. 1. BrdU pulse labelling revealed that DNA replication occurs in harvested and fixed with PFA. Incorporated CldU and IdU as well as lamin B micronuclei. HeLa cells (A–G) or COLO 320DM cells (H) were pulse treated protein were simultaneously detected, and DNA was counterstained by with 100 lM BrdU for 2 h and then harvested and fixed with PFA. The 4#,6-diamidino-2-phenylindole (DAPI). The representative images are shown location of BrdU incorporation and the lamin B protein was visualised in B–G. Mononucleated cells that were CldU-labelled, but not IdU-labelled simultaneously, and confocal images were obtained (A–F). Micronuclei were (D–F), were designated as G2 phase cells. The COLO 320DM and the HeLa classified according to lamin B status and BrdU incorporation. After culture contained 5.5 and 32.5% G2 phase cells, respectively. The number of examination of 149 (HeLa) or 115 (COLO 320DM) micronuclei, each type of micronuclei in G2 phase cells, as classified by the presence or absence of CldU micronuclei were quantified and plotted in G and H. signal and lamin B within micronuclei, were plotted (H and I).

324 DNA replication in micronuclei

shown). However, labelling the cells with BrdU for an entire cell cycle (24 h) resulted in the delay of cell cycle progression, perhaps due to the cytotoxic effect of BrdU. Furthermore, continuous BrdU labelling induced DNA damage in the labelled chromatin, which generated BrdU-positive micronuclei after that hindered analysis. To overcome this technical obstacle, we used a double labelling technique in the following section. DNA replication occurs in all lamin B-positive micronuclei but not in lamin B-negative micronuclei To determine whether every micronucleus is replicated during Sphase(10 h), we performed a double pulse-chase experiment using IdU and CldU, instead of a single BrdU pulse labelling. Namely, we cultured logarithmically growing COLO 320DM Downloaded from https://academic.oup.com/mutage/article/27/3/323/1055360 by guest on 28 September 2021 cells in the presence of CldU and cytochalasin B for 12 h, added IdU during the last 20 min of culture (Figure 2A) and then harvested the cells. We fixed the cells with PFA and detected the incorporated CldU, IdU and lamin B protein simultaneously by using specific antibodies. CldU or IdU, as BrdU, was a halogenated thymidine analogue that label the site of DNA replication. Its labelling pattern among the nucleus correlates to the position in the S phase, as shown above for BrdU-labelling. When cells in late S phase were harvested, the IdU label showed a late S phase pattern and the CldU labelling represented DNA in the main nucleus that was replicated during the 12 h prior to harvest (Figure 2B). If the cells passed through mitosis, they appeared as binucleated cells since cytochalasin B inhibited (Figure 2C). Therefore, G2 phase cells were identified as cells containing a CldU signal in the absence of an IdU signal. The representative image Fig. 3. Simultaneous imaging of DNA replication sites, lamin B and DMs shown in Figure 2D shows that a CldU signal was detected revealed the timing of replication in DM-type and chromosome-type in the lamin B-positive micronucleus. Out of the 85 lamin B- micronuclei. Logarithmically growing COLO 320DM cells were treated with positive micronuclei (Figure 2H), 84 were labelled with CldU, EdU for 2 h and fixed with PFA. The incorporated EdU, lamin B and DMs were simultaneously detected by different fluorescent colours, and DNA was while only one was not (Figure 2F). It is possible that S phase counterstained with 4#,6-diamidino-2-phenylindole (DAPI). The representative progression was partly delayed in that the cell lacking a CldU images are shown in A–E. COLO 320DM (F) or HeLa (G) cells were cultured signal and thus required .12 h of CldU labelling. Apart from in the presence of 4 lg/ml of cytochalasin B for 24 h. The cells were treated this exception, we concluded that all lamin B-positive micro- with 100 lM BrdU for 30 min and then fixed with PFA. BrdU was detected as in above and counterstained with DAPI. The binucleated cells showing nuclei are replicated during S phase. By contrast, lamin B- differential replication timing are shown. negative micronuclei in G2 phase were not labelled with CldU (Figure 2E and H), suggesting that DNA replication does not occur in lamin B-negative micronuclei. Similar results were pattern of BrdU labelling (5). DNA replication takes place at obtained for cell cycle-synchronised HeLa cells (Figure 2I). the inner euchromatin during early S phase (Figure 1B, C, D and F), at the peri-nuclear and the peri-nucleolar heterochro- DNA replication timing of DM-type and chromosome-type matin during middle S phase (Figure 1A and E) and at the micronuclei internal large heterochromatin during late S phase (not shown During the above experiment, we noticed that some micro- in Figure 1 and see Figure 2B). We counted micronuclei with/ nuclei actively incorporated IdU despite the fact that its without lamin B and with/without BrdU labelling at each cell neighbouring nucleus had already completed DNA synthesis cycle stage (Figure 1G and H) and found that no lamin B- (Figure 2G). Therefore, we examined the replication timing negative micronuclei displayed a BrdU signal at any point of not only the usual chromosome-type micronuclei but also within the cell cycle. On the other hand, we were able to detect DM-type micronuclei in COLO 320DM cells. DMs are extra- DNA replication in lamin B-positive micronuclei, most fre- chromosomal, transcriptionally active euchromatin (16,17)that quently during early S phase. However, even during early S, the are replicated during early S phase within the nucleus (18). To BrdU signal was detected in only a portion of the lamin B- detect DMs, lamin B and the DNA replication sites simul- positive micronuclei. There are two possible explanations for taneously, we labelled DNA replication sites with EdU. We this observation (i) DNA replication is permitted in only a used EdU instead of halogenated thymidine analogue, because portion of lamin B-positive micronuclei or (ii) DNA replication detection of EdU required only chemical reaction, and it enabled is permitted in all lamin B-positive micronuclei, but the limited the simultaneous detection of lamin B and DMs. The represen- BrdU labelling time (2 h) may prevent all lamin B-positive tative images are shown in Figure 3A–E, and all the images micronuclei from being labelled since S phase typically spans contain lamin B-positive DM-type micronuclei. In Figure 3A 10 h. We found that the number of BrdU-positive micronuclei and B, DNA replication was detected in the micronucleus when increased when the BrdU labelling time increased (data not the nucleus was at early S phase (A) or middle S phase (B). As

325 A. Okamoto et al. shown in Figure 3C, DNA replication was not detected in the that the resulting chromosome fragments will segregate micronucleus, when the nucleus was in late S phase. Strikingly, properly to the daughter cells after the cell division. Thus, there were other micronuclei that replicated while the nucleus the next task will be to determine whether the sister chromatids was in late S phase (Figure 3D) and in G1/G2 (no nuclear EdU- generated in micronuclei are segregated and distributed equally labelling; Figure 3E). These events were quantified, and the to the daughter cells during the next mitosis. However, this frequencies were plotted in Figure 4A. The micronuclei con- may be difficult to determine since we previously showed that taining EdU signals were most frequently associated with nuclei cells containing micronuclei frequently produced cells with in early or middle S phase. Because DMs are early replicating additional micronuclei after undergoing mitosis (6). euchromatin in the nucleus (18), the results suggest that the By contrast, DNA replication was not detected in lamin- timing of replication in the nucleus was similar to the one in negative micronuclei, indicating that their chromatin may be the micronuclei. However, a portion of DM-type micronuclei lost during cell division. These lamin-negative micronuclei replicated while the nucleus was in late S to G1/G2 phase were likely generated from broken chromatin bridges during (Figures 3D and E and 4A), consistent with Figure 2G, anaphase (6) or from interphase nuclear budding through the suggesting that the replication timing of some DMs is not lamina break (Koh-ichi Utani, Atsushi Okamoto and Noriaki Downloaded from https://academic.oup.com/mutage/article/27/3/323/1055360 by guest on 28 September 2021 synchronised with nuclear replication. On the other hand, the Shimizu, submitted for publication). Notably, transcription was replication timing of the chromosome-type micronuclei ap- not detected inside these micronuclei (16). We reported that the peared to be distributed more randomly because approximately nuclear localisation signal-bearing protein did not enter the half of the micronuclei replicated at each time point examined lamin-negative micronuclei (16). Therefore, the protein re- (Figure 4B). quired for replication and transcription may be missing in these micronuclei. Understanding the nature of DNA replication and Differentiation of replication timing between the micronucleus segregation within micronuclei will be important for de- and the nucleus termining how the entrapment of genetic material within inert As described above, we demonstrated that the timing of micronuclei may contribute to the malignant nature of human micronuclei replication differed from that of nuclear replica- cancer cells. tion, despite being within the same cytoplasm. Interestingly, we also found that the timing of DNA replication also differed between two daughter nuclei within the same cytoplasm, i.e. Funding binucleated cells, generated by cytochalasin B treatment This work was supported in part by a Grant-in-Aid for (Figure 3F and G). When 180 binucleated cells that contained Scientific Research (B) (17370002) and a Grant-in-Aid for at least one nucleus in S phase were examined, we found that Challenging Exploratory Research (21657051) both from the the replication timing was identical between sister nuclei in 162 Japan Society for the Promotion of Science to N.S. and cells but was different in 18. A similar phenomenon has been a Grant-in-Aid for Scientific Research on Priority Areas— reported previously (19,20). Thus, it appears that the difference Nuclear dynamics (19038016) from the Ministry of Education, in replication timing between the micronucleus and the nucleus Science, Sports and Culture of Japan to N.S. parallels the difference in replication timing between the two nuclei of binucleated cells. Importantly, the time when DMs are replicated may differ from the time when the same DMs are Acknowledgements replicated in the nucleus, i.e. early S phase. This difference in Conflict of interest statement: None declared. replication timing may affect transcriptional control and may ultimately influence tumour cell phenotype. References Implications of this study 1. Heddle, J. A., Fenech, M., Hayashi, M. and MacGregor, J. T. (2011) Here, we showed for the first time that lamin B-positive Reflections on the development of micronucleus assays. Mutagenesis, 26, 3–10. micronuclei replicate during the cell cycle and likely prevents 2. Fenech, M., Kirsch-Volders, M., Natarajan, A. T. et al. (2011) Molecular the dilution of micronuclear content during cell division. DNA mechanisms of micronucleus, nucleoplasmic bridge and nuclear bud replication within a micronucleus does not necessarily mean formation in mammalian and human cells. Mutagenesis, 26, 125–132. 3. Shimizu, N., Misaka, N. and Utani, K. (2007) Nonselective DNA damage induced by a replication inhibitor results in the selective elimination of extrachromosomal double minutes from human cancer cells. Genes Cancer, 46, 865–874. 4. Shimizu, N. (2011) Molecular mechanisms of the origin of micronuclei from extrachromosomal elements. Mutagenesis, 26, 119–123. 5. Shimizu, N., Itoh, N., Utiyama, H. and Wahl, G. (1998) Selective entrapment of extrachromosomally amplified DNA by nuclear budding and micronucleation during S phase. J. Cell Biol., 140, 1307–1320. 6. Utani, K., Kohno, Y., Okamoto, A. and Shimizu, N. (2010) Emergence of micronuclei and their effects on the fate of cells under replication stress. Plos One, 5, e10089. 7. Terradas, M., Martin, M., Tusell, L. and Genesca, A. (2010) Genetic activities in micronuclei: is the DNA entrapped in micronuclei lost for the Fig. 4. DNA replication timing was conserved in micronuclei, but at times cell? Mutat. Res., 705, 60–67. differed significantly from the timing in the nucleus. Slides prepared similarly 8. Shimizu, N., Shimura, T. and Tanaka, T. (2000) Selective elimination of to slides from Figure 3A–E were examined. The cell cycle stage of the nucleus acentric double minutes from cancer cells through the extrusion of was estimated by the EdU labelling pattern. Of the cells in the culture, 5.5% micronuclei. Mutat. Res., 448, 81–90. contained micronuclei. A total of 109 micronuclei with DMs (A; DM type) and 9. Kramer, J., Schaich-Walch, G. and Nusse, M. (1990) DNA synthesis in 114 micronuclei without DMs (B; chromosome type) were classified according radiation-induced micronuclei studied by bromodeoxyuridine (BrdUrd) to the presence or absence of lamin and EdU signal within micronuclei. labelling and anti-BrdUrd antibodies. Mutagenesis, 5, 491–495.

326 DNA replication in micronuclei

10. Tanaka, T. and Shimizu, N. (2000) Induced detachment of acentric chromatin from mitotic chromosomes leads to their cytoplasmic localiza- tion at G1 and the micronucleation by lamin reorganization at S phase. J. Cell Sci., 113, 697–707. 11. Shimizu, N., Kanda, T. and Wahl, G. M. (1996) Selective capture of acentric fragments by micronuclei provides a rapid method for purifying extrachromosomally amplified DNA. Nat. Genet., 12, 65–71. 12. Shimizu, N. and Shingaki, K. (2004) Macroscopic folding and replication of the homogeneously region in late S phase leads to the appearance of replication bands in mitotic chromosomes. J. Cell Sci., 117, 5303–5312. 13. Salic, A. and Mitchison, T. J. (2008) A chemical method for fast and sensitive detection of DNA synthesis in vivo. Proc. Natl Acad. Sci. U S A, 105, 2415–2420. 14. Willingale-Theune, J., Schweiger, M., Hirsch-Kauffmann, M., Meek, A. E., Paulin-Levasseur, M. and Traub, P. (1989) Ultrastructure of Fanconi anemia fibroblasts. J. Cell Sci., 93 (Pt 4), 651–665. Downloaded from https://academic.oup.com/mutage/article/27/3/323/1055360 by guest on 28 September 2021 15. Paulin, L. M., Blake, D. L., Julien, M. and Rouleau, L. (1996) The MAN antigens are non-lamin constituents of the nuclear lamina in vertebrate cells. Chromosoma, 104, 367–379. 16. Utani, K., Kawamoto, J. K. and Shimizu, N. (2007) Micronuclei bearing acentric extrachromosomal chromatin are transcriptionally competent and may perturb the cancer cell phenotype. Mol. Cancer Res., 5, 695–704. 17. Shimizu, N., Hanada, N., Utani, K. and Sekiguchi, N. (2007) Inter- conversion of intra- and extra-chromosomal sites of gene amplification by modulation of gene expression and DNA methylation. J. Cell. Biochem., 102, 515–529. 18. Shimizu, N., Ochi, T. and Itonaga, K. (2001) Replication timing of amplified genetic regions relates to intranuclear localization but not to genetic activity or G/R band. Exp. Cell Res., 268, 201–210. 19. Sandberg, A. A., Sofuni, T., Takagi, N. and Moore, G. E. (1966) Chronology and pattern of human chromosome replication, IV. Autora- diographic studies of binucleate cells. Proc. Natl Acad. Sci. U S A, 56, 105–110. 20. Takanari, H., Yamanaka, H., Mitani, H. and Izutsu, K. (1994) Replication sites as revealed by double label immunofluorescence against proliferating cell nuclear antigen (PCNA) and bromodeoxyuridine (BrdU) in synchro- nized CHO cells and vincristine-induced cells. Biol. Cell, 82, 23–31.

327